Patients with relapsed acute myelogenous leukemia (AML) have a uniformly grim prognosis; only 13-40% achieves a complete response (CR/CRi) with currently available salvage chemotherapy regimens. Long-term survival can only be achieved by those who proceed to an allogeneic cell transplant after achieving a CR2. In spite of advances in our understanding of the biology of both AML and relapsed AML, there have been no significant improvements in either treatments or outcomes in the past 10-20 years. This suggests the need for innovative new approaches to overcome inherent biologic resistance of relapsed AML. We hypothesize that targeting the interaction of AML with the bone marrow (BM) microenvironment will overcome the acquired resistance factors inherent in relapsed AML and result in improved CR rates and overall outcomes. We and others have shown that the BM microenvironment provides an important protective effect against genotoxic stresses such as chemotherapy and that physical interruption of this interaction render AML cells sensitive to chemotherapy in vitro and in vivo. We propose to target the CXCR4-SDF-1 axis using a small molecule bicyclam, AMD3100 in conjunction with Granulocyte Colony Stimulating Factor (G-CSF), to promote rapid and sustained release of AML blasts from the BM microenvironment in patients with relapsed AML. In Aim 1 we will perform a phase I/II clinical trial in patients with relapsed AML. G-CSF+AMD3100 will be administered prior to mitoxantrone-etoposide-high dose Ara-C (MEC) salvage chemotherapy to optimally mobilize leukemic blasts, thus sensitizing the AML blasts to MEC chemotherapy. We will assess the safety and toxicities of G- CSF+AMD3100, their impact on multi-lineage hematopoietic recovery and, most importantly, on CR rates. In Aim 2 we will measure the magnitude of AML mobilization after G-CSF +AMD3100 administration and attempt to characterize the phenotype and cell cycle status of AML blasts and primitive (CD34+/CD38-) subsets of AML before and after G-CSF+AMD3100 treatment. In addition we will functionally assess the impact of G- CSF+AMD3100 on mobilizing Leukemia Stem Cells (LSCs) using in vitro LTC-IC and in vivo NOD-SCID repopulating cell (SRC) assays. In Aim 3 we will use a novel and informative models of mouse AML and human ALL to explore the role of small molecule inhibitors of two other critical axes (VLA-4/VCAM-1 and Selectin/Selectin Ligand) on normal and leukemia stem cell mobilization and homing to the BM. The effects of these small molecules will be tested alone or in combination with CXCR4 blocking agents (AMD3100, 070, G- CSF and ALT1188). VLA-4 and pan-selectin inhibitors will be used in these models alone or in combination to determine if they also can be used to sensitize leukemic cells to chemotherapy in vivo. These studies will provide insights on the effect of disrupting AML-BM microenvironment interactions using AMD3100, G-CSF and other small molecules in the future. This approach may represent a simple and non-toxic way of increasing chemotherapy sensitivity and overcoming inherent genetic and epigenetic resistance factors associated with AML thus improving outcomes of patients with relapsed and resistant AML.